Extra teros ports in synoptic W plots log inconsistently

[stephpenn1]

Sensor ports 3 and 4 in synoptic wetland sites report as either NAN or 0 but are not consistent. We'd prefer these to read in as NAN if there is no sensor install, but aren't sure why we see both?

[bpbond]

Checked the design table and we're handling this correctly as far I can see; this is a question for @roylrich and the datalogger code. I am removing from our v1-2 milestone.